REPORT 1307 LIFT AND CENTER OF PRESSURE OF WING-BODY-TAIL COMBINATIONS AT SUBSONIC, TRAN- SONIC, AND SUPERSONIC SPEEDS By Wauuant 0. Prove, Jaox N. Nrsraux, and Guonon B, Kaareant SUMMARY A method is presented for caloulating the lift and. center-of- pressure characteristics of cireular-cylindrical bodies in com- Bination with triangular, rectangular, or trapezoidal winge or tails through the subsonic, transonic, and supersonic speed ranges. The method is restricted to wings which are unbanked cand do not have sweptback trailing edges or sweptforward leading edges. The method is further restricted to small angles of at- tack and small angles of wing and tail incidence. To obtain the wing-body interference, certain factors are defined that are the ratios of the lift on the components in combination to the lift on the wing alone. These ratios are obtained primarily by elen- der-body theory. The wing-tail interference is treated by assum- ‘ing one completely rolled-up tortes per wing panel and evaluating the tail load by strip theory. A numerical example is included to show that the computing form and design charts presented reduce the caleulations to routine operations. Comparison is made between the estimated and experimental characteristics for a large number of wing-body and wing-body-tail combina tions, Generally speaking, the lifts were estimated to within £10 percent and the centers of pressure were estimated to twithin 0.02 of the body length. INTRODUCTION ‘The problems of the interference among the components of airplanes or missiles have received much attention be- cause of their great importance in high-speed aircraft design, ‘This importance is due to the interest in designs employing largo fuselago radii and tail spans relative to the wing span. One of the notable methods for determining wing-body interference at subsonic speeds is thet of Len- nertz, reference 1; data supporting the work of Lennertz are presented in reference 2. _Laborious methods are avail- ablo (refs. 3, 4, and 5) for computing the interference load distributions of wing-body (or tail-body) combinations at supersonic speeds, A simple method is presented in refer- ence 6 for estimating the effects of wing-body interference ‘on lift and pitching moment when the wing is triangular. One of the notable methods for calculating wing-tail inter- ference in subsonic aircraft design is that of Silverstein and Katzoft in references 7 and 8. For supersonic speeds, ‘Morikawa (ref. 9) has examined the four limiting cases of zero and infinite aspect ratio for wing and tail and hes found that the loss of lift due to interference can be as large as the lift of the wing itself for equal wing and tail spans. 00101087 Using slender-body theory, Lomax and Byrd (ref. 10) have analyzed the wing-tail interference of a family of combina- tions having swept wings. Several authors have studied problems of the nonuniform downwash field behind wings in combination with a body at supersonic speeds; Lager- strom and Graham (ref. 11) present solutions for certain vortex models representing the downwash field, ‘The assumption of one fully rolled-up vortex per wing panel should provide a good prediction of the downwash even relatively close behind unbanked low-ospect-ratio triangular swings at small angles of attack. However, for large aspect ratios or high angles of attack more than one vortex per swing panel is probably needed to provide agreement between theory and experiment. With regard to the problem of determining the tail loads due to a nonuniform downwash field, Lagerstrom and Graham (ref. 11) advocate the use of strip theory. Alden and Schindel (ref. 12) have developed a method based on linear theory for determining the tail Joad in certain eases. ‘Tho purpose of the present report is twofold: first, to pre~ sent a unified procedure for calculating interference effects and to examine the assumptions underlying the procedure; and, second, to compare the predictions of the method with experiment in order to estimate the accuracy of the predic- tions and their rango of application. syMBoLs: Ar tail-alone aspect ratio Ar wing-alone aspect ratio a mean serodynamie chord of wing alone or tail alone, in, ¢ chord at wing-body juncture or tail-body juncture, in, ¢ tip chord of wing’or tail, in. ¢ wing chord at spanwiso distance y from body axis, ©, —__hinge-moment coefficient based on wing-alone aren ©, rate of change of hinge-moment coeflicient. with angle of attack, per radian Oy, rate of change of hinge-moment coefficient with ‘wing incidence angle, per radian "tb wing oo tle flare deed obo tbe eed ys of to ager tained toni 867568. Lr Law M Me REPORT 1507-—NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS life coefficient based on wing-alone area except tail- lone lift coefficiont based on tail-slone area lift-eurve slope for angle of atteck, per radian (un- less otherwise specified) lift-curve slope for wing or teil incidence, per radian (unless otherwise specified) pitchingmoment coefficient. based on wing-alone area pitching-moment-curve slope for angle of attack, per radian (unless otherwise specified) pliching-moment-curve slope for wing-incidence angle, per deg body diameter, i complete elliptic integral of second kind ‘wing vortex semispan at tail position, in, wing vortex semispan at wing trailing edge, in. swing vortex semispan for large downstream dis- ‘tances, in. Alden-Schindel influence coefficient at spanwise distance 7 image vortex semispan at tail position, in. image vortex semispan at wing trailing edge, in. height of wing vortex above body axis at tail center of pressure, in. tail interference factor ratio of lift component to lift of wing alone or teil ‘alone for variable wing or tail incidence ratio of lift component to lift of wing alone or tail alone for variable angle of attack ratio of lift of body nose to lift of wing alone length of wing-body-tail combination, in distance from most forward point of body to inter section of wing leading edgo and body, in. distance from most forward point of body to center of moments, in. moment reference length, in. distance fom most forward point of body to shoulder of body nose, in distance from most forward point of body to inter- section of tail leading edge and body, in. distance from most forward point of body to center of pressure position, in, lift force, 1b lift on tail section duo to wing vortices, Ib life on body seotion between. wing end tail due to swing vortices, Ib cotangent of leading-edge eweep angle pitching moment, Ib-in. free-atream Mech number slatic pressure difference between top and bottom of wing, Ibjsq in. free-stzeam dynamic pressure, Ib/sq in. ody radius, in. body radius at shoulder of nose, in- ody radins at wing, in. body radius at tail, in. Reynolds number based on @ of larger lifting surface 2 he Aue Pe "g T g Basszs908 ‘maximum semispan of wing or tail in combination with body, in cross-sectional area of nose at maximum section, sq in. reference area of combination lift coefficient, sq tail-alone area, 9q in, swing-alone area, sq in. ratio of wing maximum thickness to chord Iength volume of body, considering the body as eylindrioal behind the position of maximum cross section, volume of body nose up to shoulder, eu in. free-stream velocity, in,/see streamwise, spanwise, and vertical coordinates, respectively distance to center of pressure measured from inter- section of wing leading edge and body for wing quantities and from intersection of tail leading edge and body for tail quantities, in, distance to local conter of pressure at spanwise distance y measured from intersection of wing leading edge and body, in. distance from intersection of wing leading edge and body to wing hinge line, i angle of attack of body centerline or of wing ‘alone, radian (unless otherwise specified) local angle of attack at spanwise location y from body axis, radians IT wwing-alone or tail-alone effective aspect ratio circulation, positive counterclockwise facing up- stream, sq in,/see circulation at wing-body juncture of combination, 2q in, foc wing-or tail-incidence angle, radians wing semiapex angle, deg, spanwise varieble of integration taper ratio, (%) sweep angle of leading edge, dog sweep angle of trailing edge, deg free-stream density, slugs/ou in. body ‘combination, either wing-body or wing-body-tail combination minus nose forobody body nose tail swing vortex wing Alden-Schindel theory body in presence of tail ody in presence of wing strip theory tail in presenco of bodyLIFT AND CENTBR OF PRESSURE OF WING-BODY-TAIL COMBINATIONS W(B) wing in presenco of body @ variable, 8 constant a 8 variable, «constant W(B)a_ wing in presence of body and « variable, 5 constant Other compound subscripts to be interproted similarly to tho preceding compound subscript. GENERAL THEORETICAL CONSIDERATIONS Before presenting the detailed development of the method, tan outline of the appronch to be followed is presented, ‘The theory is restricted to small angles of attack and small angles of wing and tail incidence. Attention is focused on pointed bodies having wings and tails mounted on body sections of uniform diameter. For the sako of consistency, the forward lifting surfaces are termed the wings, even in cases of canard configurations. Both wings and tails may have variable inci- dence, but cases of differential incidence between opposite panels of the wing or tail are beyond the scope of this report. Teil efterbody Weg 7 + Noe. Wing —=}— oersody fT sony o | 1 1 1 . I | 1 I io ff 1 1 t | " cre > tam > Dean S f { I~. 7 I ! =o o tod 1 i le 1 ! ieee i i 1 1 | i ot } | i @ (8) Parts of a wing-body-tail combination, () Lifts without wing-tail interference. (@) Lifts due to wing vortices. Fravmn 1.—Parts and lift components of a wing-body-teil eombination ‘Tho terminology is indicated in figure 1 (e). ‘The nose is that part of the body in front of the wing. However, when the wing is mounted on an expanding section of the body, the nose is taken to be the entire expanding part of the body. For the purpose of analysis, the lift of the wing- body-tail combination is taken to be the sum of the seven 569 principal components indicated in parts (b) and (¢) of figure 1" ‘These components are: . Lift on nose ineluding forebody, Lay . Lift on wing in presence of body, La Lift on body due to wing, Zsar) . Lift on tail in presence of body, Lrw) 5. Lift on body due to tail, Leen Lift on tail duo to wing vortices, Lx) . Lift on wing afterbody due to wing vortices, Lac) All eoefiicients, except those for the tail alono, aro based on tho exposed wing area. Tho lift and conter-of-pressure position calculation procedures for tail-body interference are identical to those for wing-body interference, except for a term to refer the tail-body interference lifts to the wing area; therefore, they will not be treated separately. ‘The method presented for computing the wing-body and tail-body interference (components 2 through 5) is based primarily on slender-body theory (ref. 13). In this theory, Spreiter has shown that the first term of the wave equation for the velocity potential (M.4— 1) ¢2— Pn Pu=0 a) can be ignored for slender wing-body combinations, so that equation (1) reduces to Laplace’s equation in the y,¢ plane. Using this simplification, simple, closed expressions are obtained for lift-eurvo slopes. It is well known that for wing-body combinations which aro not slonder, lift-curve slopes are overestimated by slender-body theory (ref. 6). However, this fact does not prechide the uso of slender-body theory for nonslender con- figurations sinco, in certain instances, the ratio of the lift of the wing-body combination to that of tho wing alono can be accurately predicted by slender-body theory, even though the magnitude of the lift-curve slope might be incorrect. ‘Brom the foregoing ratio, which is called Ko, and a good estimate of the wing-alone lift-curve slope, the lift-curve slope of the combination can be obtained. ‘This was essen- tially the method used by Nielsen, Katzen, and Tang in reference 6 to predict the lift and moment characteristics of ‘triengular wing-body combinations. Good agreement be- tween experiment and theory was obtained. With these facts in mind, the method used by Morikawa (ref. 14) for presenting lift interferenes is adopted. In this method, the wing alone is defined as the exposed half-wings joined together. The lift of the combination is related to the lift of the wing alone by the factor Ke which is to be determined. Le=Kelw @) The factor Ke is decomposed into three factors Kan, Kray, and Ky Which represent tho ratios of tho body lift wing lift, and nose lift of the combination to that of the wing alone. Kom Koonyt Kwan Kei ®) (Gs) 2 Kane ES ® (Coan Kn Gen “Go. b=0 (6)570 Iy_(Ciadn 4 Te Coole . ‘The factors Kyo) and Ky are defined for the case in which the angle of attack of the combination is varying but the wing- (or tail) incidence angle is zero. For the caso in which the incidence angle is varying but the angle of attack of the body is zero, two analogous factors are defined. ax @ a=0 ®) So far, only & way of representing lift results has been presented. The solution of a problem requires a determina- tion of each of these ratios. ‘Then, the lift on any component can be estimated from the wing-alone lift-curve slope. The best value of the wing-alone lift-curve slope that is available should be used; preferably the experimental value. The detailed determination of each of theso ratios is presented in subsequent sections of this report. In general, slender- body-theory values are computed. ‘These are compared with values computed by other methods and ultimately with experimental results. ‘There are some conditions for which slender-body theory is invalid or for which more exact methods are available. ‘Theso are pointed out and tho slender-body-theory values for the ratios are replaced. LIFT THEORY ‘Tho lift theory as developed is for the angle-of-atiack range over which tho lift curves are linear and is equally applicable to subsonic end supersonic speeds unless otherwise noted. > upr ox popy nose ‘From equation (6) Ciy=Eer (Cra) @) For the caloulations in this report, Zy is evaluated by use of slender-body theory, Bann io) Ty=0 a1) so that Soe ae © It is known that slender-body theory is usually not suili- ciently accurate to determine body-alone lifts in cases such ‘as nonslender bodies or large angles of attack. However, for combinations which are not predominantly body, the nose lift is not a large part of the total lift, and slender-body theory generally gives satisfactory results. ‘If improved accu- racy is desired, linear theory, the viscous cross-flow theory of reference 16, or experimental results ean be used. REPORT 1307—NATIONAL ADVISORY COMMITTED YOR AERONAUTICS LWT ON WING IN PRESENCE OF BODY Angle of attack.—From equation (5) Cary c= = Kr oy (Clg) w® as) when 5=0. The value of (Cre)y from experiment should e used if available; otherwise the value from linear theory should be used. Therefore, obtaining Cry depends on obtaining Keren. ‘The value of Kirin given by slender-body theory (ref. 14) Evo= {Comrie De4 Gy (14) (Tho assumption is made that no negative lift is developed behind the maximum wing span. Jones (ref. 16) has pointed out that for wings, at least, the negative lift predicted on ‘theso sections by’ slender-body theory is prevented by separation.) ‘This function is plotted in chart 1. In the limiting case of r/s=0 the combination is all wing and the value of Kray=1. As r/s approaches unity, there is a very: small exposed wing. For this small wing, tho body is effectively e vertical reflection plane and the angle of attack is 2a due to upwash (as is discussed later). ‘This makes Evy Te is clear that the values of Kis should be satisfactory for slender wing-body combinations. However, they eannot bo used for large aspect ratios, for which slender-body theory is inapplicable, without further investigation. An approxi- mate method for evaluating Kwrm is to suppose that the exposed wings are operating in the upwash field of the body alone and then'to calculate the resultant wing lift, Neglect ing any effect of tho noso, it has been pointed out (ref. 17) ‘that the upflow angle due to the body varies spanwise on the horizontal plane of symmetry as aymay(1 +5) (15) where y is the Interal distance from the body axis. ‘The wing is thus effectively twisted by the body-alone flow. If now tho upwash angle given by equation (16) is taken into account by using strip theory, an approximate value of Ey is obtained as follows: 7 I aa (16) Equation (16) does not include tip effects. The following expression is obtained in terms of r/e and taper for wings of uniform taper. an It is notable that Kyi») doos not depend on aspeot ratio.LIFT AND CENTER OF PRESSURE OF WING-BODY-TAIL COMBINATIONS Equation (17) was used to determine Kyra for \=0, 4, and 1, and these results are compared to those of slender body theory in figure 2. Tt is seen that the effect of taper is small compared to the effect of r/e. Both theories give nearly the sume values at both high and low r/a, but the upwash-theory values are, in all instances, greater than those of slender-body theory. Nowhero is the difference of great significance, Although account has been taken of the upwash induced along the wing span by the body in the determination of Kw by upwash theory, no account has ‘eon taken of the loss of lift due to interaction between the wing and the body of tho winged part of the combination, For this reason, Kye will be too large. Therefore, the slender-body-theory values of Ky) should be used for all combinations. Slepder= 6 | % = = © 7 0 Bedy-rodus,wing-semspon rota, 5 Frounn 2.—Comparison of Kya) or Kran determined by siender-body ‘and upwash theories, b7l For wing end body combinations with lange-aspect-ratio rectangular wings the linear-theory solution for Kyi) is available (ref. 18). ‘Theso results aro presented in chart 2 where they are compared with the slender-body-theory results, Since a graphical integration was required for the determination of the linear-theory values, there is a small uncertainty in th result, represented by the cross-hatched area, For a fixed value of r/s and for the range 21, then ky from linear theory should be used. Tt might be surmised that the present method of deter mining the lift on wing in the presence of the body is applicable at subsonic speeds since the slender-body-theory values of Kyi and kya on which itis based are not depend- ent on Mach number and the effect of Mach number enters only through (G:,),. ‘This supposition is subsequently borne out by experimental date. Spreiter mado the observa~572 tion in reforenco 18 that tho loading on the minimum drag wing-body combination of Lenneriz,(rof. 1) is identical at low speeds to that of a slender wing-body combination with ‘a body of uniform diameter. ‘The division of lift between ‘wing and body based on this loeding is shown in figure 3. Since the present method is based on the division of lift as given by Spreiter, the equality of the results of Spreiter and Lennertz is further evidence of the applicability of the prosent method to subsonic speeds. At this point, it is desirable to consider the effects of span loading on the division of lift between wing and body because this information has bearing on tho validity of the vortex ‘model used in determining some later results. Besides his result for minimum drag, Lennertz also determined the division of load between wing and body for uniform span loading. This result, which corresponds to replacing each sido of the combination by a horseshoe vorter, is shown in figure 3, wherein the part of the lift carried by the body is shown as a function of the ratio of body radius to vortex semispan. For the samo valuo of the abscissa there is not much difference between the fractions of the lift acting on the body for the two cases. Generally, the span of a horse- shoo vortex replacing a wing is less than the wing span. If ‘account is taken of this fact in the comparison, the existing difference would largely disappear. Thus, the represente- tion of the wing-body combination by a horseshoe vortex on each side is compatible with the present method of determining the division of lift between wing and body. Angle of attack —From equation (4) Gogg =Kna (Cl) w ® ). The slender-body theory value of Kaor) is (HO (20) Ean ‘This function is plotted in chart 1. In the limiting caso of r/e=0 the combination is all wing and Kygm=0. As r/s ap- proaches unity, there is a-very small exposed wing. For this ‘small wing the lift on the body due to the wing is the samo as the lift on the wing itself. Thus, Kaan—Kiriy=2. ‘To determine the applicability of the slender-body-theory values of Kyqn to nonslender combinations, Knqr is now determined by an independent method. On the basis of slender-body theory, nonexpanding sections of a body in a uniform flow develop no lift. Therefore, the lift on = straight portion of a body on which a wing is mounted is duo principally to lift transmitted from the wing to the body. ‘A point on the wing is thought of as a souree of lifting dis- turbances which move in all directions in the downstream Mach cono from the point. Some of these disturbances traverse the body. ‘The assumption is made that the solo effect of the body (regardless of cross section) is to displace these pulses downstream without diminishing their lifting potential. ‘This is the so-called delayed reaction of Lager- REPORT 1307—-NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS 40 Soa Lang iin Ee, Sen damn he & / Bo 8 AAR torn eosen 5 tert 4 7 oy 1 2 3 Redivs-semispon rato, (hy ore Frovus 8.—Comparison of slonder-body theory and theory of Lone nert for fraction of if cared by body. (= sdf} 3 strom and Van Dyke in reference 19, which was substantiated for a particular family of rectangular wing-body combinations in reference 3. Downstream of the wing, the flow returns to tho free-stream direction. ‘The offect of this change in flow direction is felt on the surface of tho afterbody behind tho Mach helix originating at the trailing-odge, root-chord juncture, In this region, the reaction tends to cancel the lift transmitted from the wing onto the body. ‘The effective resultant lifting area on the body for one half-wing can thus >be approximated by the shaded area shown in figure 4(a). ‘While a nonplanar model has been sot up to ropresont tho lift transmitted to the body from the wing, further simplificn~ tion to an equivalent planar case is desirable before calcula- tions are performed. ‘The body is imagined now to be col- lapsed to @ plane and the Mach helices of figure 4(a) become tho Mach lines of figure 4(b). ‘The lifting area of tho body is tho shaded aroa of figure 4(b) which is at zero anglo of attack. ‘This area is equal to the horizontal projection of the lifting area of the actual body surface (fig. 4(a)). ‘The lift on 3 (21)LIFT AND CENTER OF PRESSURE OF WING-BODY-TAIL COMBINATIONS 1 poet Regen of influence of ‘ing or toll en body é och ines—— (0) Nocploar Node (8) anor Mode Frounn 4,—Equivalent planer model for determination of Kaun and Kaen for high-aspect-ratlo range at supersonic speeds. 573 the body can be calculated simply by integrating pressures duo to the half-wing over the shaded are and doubling the result, In determining the pressure field of tho half-wing on the planar area, both subsonic and supersonic leading edges are considered. ‘Tip effects are not considered, and tho analysis is confined to the case in which the Mach line emanating from the leading edgo of the wing tip falls behind tho region of lift carry-over onto the body. This condition imposes the restriction eatin) (gi+1)>4 2) ‘on the wings for which the method is to apply. ‘The valuo of lift transmitted to the body by a half-wing with a supersonie leading edge is given (using the solution of ref. 20) a8, : Evens Tanase a8 (ay [OM uP "br Bm Seno rat in terms of the coordinate system of figure 4 (b). This result is doubled to account for the lft of two half-wings and divided by tho lift of the wing alone to obtain Kn). For all super- sonic Mach numbers Kags is Kao where mB>1. reference 21, Koon= omtyttom] — [rtatemet — CE otal] 4 Goal} B14) (02)($-1) 00.) y [TF emt @m-+if 2 RPT (02) en-2(14-£6) Ogg Bin (Zeon (45) haem) fe Similarly, for subsonic leading edges there is _8gacry (Am) - obtained, using the appropriate conical lifting solution from Doon Sa Oy a nf meta 8) giving 1g [orton ontoonnl [ono non ()(E1) @.e| = 2) where mB<1. The effect of body upwash in increasing the lift of the exposed wing has not been taken into account in calculating the effect of the wing on the body. It is to be noted that Kyor, in equations (24) and (26) depends on @ number of parameters, of which four are independent. However, tho quantity Ksyn(1-+»)($-1) (BCoa)y ‘a function of only mp and ‘This quantity is amt i-+mppe presented as a function of 26r/cy for constant values of mé in chart 4 (a) which is to serve as a design chart in deter- mining Kor subject to the restriction of equation (22). For the purpose of illustrating the behavior of Kaqr, and comparing equations (24) and (26) with slender-body Enon, chart 4 (a) has been used to obtain figure 5, which presents Ksor as & funetion of A and r/s for-k=0, 1/2, and 1 and for no trailing-edge sweep. ‘The ease of h=0'cor- responds to triangular wings (fig. 5 (a)), X=1 to rectangularB74 wings (fg. 5 (b)), and A=1/2 to trapezoidal wings (ig. 5 (©). For triangular wings, the curve of Kea by the present theory for BA=0 is slightly greater than Kya as given by slender-body theory and has not been included in the figures, since for such amall values of BA slender-body theory is the more valid. Incidentally, the restriction of equation (22) is met by all triangular wings with no trailing- edge sweep. An examination of figure 5 (b) for rectangular wings shows good agreement between slender-body theory and the present theory at 6A=2, the lowest aspect ratio for which the present theory is applicable to rectangular wings. In the easo of the trapezoidal wings (fig. 5 (c)), the restriction ‘of equation (22) imposes the condition that 64>4/3. For a value of fa of 4/8 there is no appreciable difference between jen 1 7 le sender boa Phedey 7 5 34 z a 3 2 1 (o) ° y a = € 5 Bodyradusving-semispon roti, of (@) Triangular wing-body combinations. Fravan_ 6—Comparison of Kaom ot Kotz) determined by slender-body theory and present theory for wings with no tralling-edge aweap. REPORT 1307—NATIONAL ADVISORY COMMUTTER FOR AERONAUTICS slender-body Koqr and the value of Koen by tho present theory. On the basis of figures 5 (a), § (b), ond 5 (c), and since wing tip effects invalidate equations (24) and (26) for PAC+N)(e-+1) <4, the following selection rule should bo vied: If 2A(1+0)(2541) <4, uso tho slonder-ody theory Kaan and if BAC-+N)(E41)>4, use Kner from chart 4. ince rectangular and triangular wings aro vory common, and since (8Cz,),, is known in closed form for theso plan forms, specialized results can readily be obtained from pa Slender oy 2" Fair) Cc) ° 7 Zz Body-rotus, wing-semispan rati/5 () Rectangular wing-body com! Ficues Continued,LIF AND CENTER OF PRESSURE OF WING-BODY-TAIL COMBINATIONS 575 equations (24) and (26) for Kogr. For rectangular wing- 8 24, | body combinations, Kaan is ce at For triangular wing-body combinations with subsonic leading edges, Kagry is tts eh ey 1 a oay( E] oi oo (28 | 2] vanes fae EE | oy Body-redh, wing-semispon roti, r/ 4 (4) (9 Tap ing con pyar and for supersonic leading edges aoung | SPE OT [ae tow a tal I |e DEG 4 BA 4901046228376 ‘Dhe case for no afterbody behind the wing can also be calculated for the high-aspect-ratio range at supersonic speeds. Tho method for determining Kyqr) without after- body is the same as with afterbody except that the upper limit of integration in equations (23) and (25) is c- rather than ¢,+6n. Carrying out these integrations multiplying by 2 and dividing by the lift of the wing alone yields for the case of no afterbody: ron|6(Cio)w] 0+ ($1) 2 ‘mB-+S5 FA) ar al ee pH 10 (iY cor pte (Gi) RST ae BE Vink F=T cosh” Fipom>l god (30) Kno {6(Cis)w]O+D ($1) sare (EO?) VG) G4)- @ (mpy*+-mp (iy (sm+1) [oye oe ETE alms $1) (Bev fs oma, a @n ‘The restriction that $>disnotaseriousone. For ¢>Sit is lear that the lift transmitted to the body is the same as for ‘Tho value of the parameter 4 * so that Kaan is constant. Ko{6(Cr)y) 0+1)($-1) is plotted as a function of me and 26rfo, in chart 4 (b). A comparison of Ksqn as determined from chart 4 (a) with that from chart 4 (b) gives an indication of the impor- tanco of the afterbody for any particular configuration. For small values of the ratio 2A\r/e)w there is very little effect of the afterbody on Ky) but, for large values, the effect can be os large as several hundred pereent. At sub- sonio speeds no distinction is made between the afterbody and no-afterbody cases. ‘The difference between the two, which is usually small in terms of total lift at supersonic speeds, is further reduced at subsonic speeds because of the lesser tendency of lift to be carried downstream. ‘Wing-incidence angle—From equation (7) Gurgn=Haon(Cc.)ybe G2) ‘Tho only general method for determining Ison is slender- body theory. Tt has been shown in reference 22 by use of REPORT 1307—NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS reciprocal theorem that for combinations with cylindrical Yodies the following equality is valid under the assumptions of slender-body theory: kaon Koren —koven 3) ‘Pho values of Kacm a8 given by equation (33) are included in chart 1. ‘An interesting approximation that gives some insight into the interrelationships between Kaun, Kies, kaon, and Key can be made. If itis assumed that the wing transmits ‘a certain fraction of its lift to the body irrespective of whether the lift is developed by angle of attack or wing- incidence angle, sn approximate value for kan, namely, Bam, is k, wm Renae Koon 4) ‘The values of kyr) and Macy #8 determined from equations (3) and (84) do not differ by more than 0.01, o quantity that is practically indistinguishablo in chart 1.” This smnall difference is due to the difference in the forms of the load distribution on the wing for lifts due to angle of attack and wing-ineidence angle. Wing-tail interference results from downwash in the region of the tail caused by the wing vortices. ‘The problem of determining wing-tail interference breaks down into tho problems, first, of determining the number, strongths, and positions of the wing vortices at the tail and, second, of determining the reaction of the tail section to the nonuniform flow field induced by the wing vortices. ‘This component of the combination lift is the most laborious to ealculate. ‘Tho same mothod is used for subsonic and supersonic speods. Line-vortex theory is used in tho solution of tho wing-tail interference problem following tho gonoral lines of othor investigators. ‘The moda to bo used is illustrated in figures 64nd 7. This model of the wing is the samo os the Lennortz model for uniform londing previously discussed and is thus compatible with the method used here for caleulating wing- body interference. Only one trailing vortex per wing panel is considered elthough more vortices per panel could bo us to obtain greater accuracy at the expanso of greater compl cation. ‘The wing trailing vortices stream backward but undergo lateral and vertical deflections as e result of the body crossflow field and the interaction between vortices. Image vortex lines are introduced inside the body at the mage position of the trailing vortices to satisty the boundary condition for a cireular body. Sufficiontly far downstream the extemal vortices approach an asymptotic spacing. Vortex characteristics—For easo of calculation it is assumed that one fully rolled-up vortex is discharged from each wing panel. While this model simulates tho flow behind the wing pancls of many combinations, thero aro cases where it does not. As examples, some results obtained dy Spahr and Dickey in the Ames 1- by 3-foot supersonic wind tunnel are presented as the solid curves in figures 8, 9, and 10. These data were obtained by the vapor-screon tech- nique described in reference 15. Figure 8 shows that for aLIFT AND CENTER OF PRESSURE OF WING-HODY-TAIL COMBINATIONS bY —w | h. 1 ! Tol center I of pressure"y Gay Fraune 6—Vortex model used in determination of wing-tall inter ference. fr Frovne 7.—Cireulation distribution at wing trailing edge and equiva- lent horseshoe vortex. low-aspect-ratio triangular wing in combination with a body at low angles of attack, only one tip vortex is present as as- sumed, However, as the angle of attack is increased a body ‘vortex appears, and as the wing aspect ratio is increased (figs. 9 and 10) an additional vortex appears from the inboard sections of the wing. Thus, the simplified model of one vortex per wing panel is not always an adequate basis for 577 ‘ . attr postin fom enor! 6-9 | has ig rts 3 | Fovorec| = 3? L-|4_ SE : +++ Theceeicol osymptotic lateral position = (o) 0 15 ™ T = lex polh in free-streom dicection = ex path correcied for erose~or ‘nd induced etfecls eee Lae 5 E Body vero ° + s 1 z Angle of stack, o, deg (@) Lateral position of vortex. () Vertical position of vortex. ‘Fioune 8—Comparizon between theory and experiment for lateral ‘and vertical positions of wing vortex 1.8 c, behind wing of sspect ratio 28 triangular wing and body combination; Mfw=2.0,r/e=0.00. computing downwash. However, sevoral investigators have successfully applied this simplified model to the computation of tail loads. ‘These results indicate that the total tail load of each of the configurations investigated is insensitive to the details of the vortex flow although the downwash behind the wing end the spanwise distribution of tail lond are not. ‘This conjecture is substantiated in part by tho theoretical work of Morikawa, reference ®, who has calculated the tail lifts of slender wing-body-tail combinations using one fully rolled-up vortex per wing panel and using a fit vortox sheet. Only for fully rolled-up vortices in the immediate vicinity of the tail tip does any appreciable difference between the two eases occur. ‘The results of Lomax and Byrd, refer- ence 10, for a family of swept wing-body-tail combinations aro in accord with the findings of Morikawa. It was on the basis of this evidence and because of its great simplicity that the uso of one wing vortex per panel was adopted. ‘The ‘adequacy of this assumption and its range of application is subsequently determined by comparison between experiment and theory. ‘The circulation distribution at the wing trailing edgo determines the strength I, and the spanwise position fr of the vortex at the trailing edge. ‘Tho actual circulation distribution is replaced by an equivalent horseshoo vortex corresponding to the Lennertz model for uniform loading. ‘Figure 7 illustrates this model. Note that figuro 7 contains tho tacit assumption that the maximum value of the cirou-578 REPORT 1307—NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS re Tel pti tom cert Cp [i {Tresre! fol potion rm coe 6 ‘ PS aa es Inboard varien| Bs Ry B 1 voted 3 2 tt = vert o| o o| 6 oh 1 vertex pth in feesrtam decten Vertes poh comets ‘or cotton Treas 2 A re noes efi | Tie ita een age te 2 Sol === vores ptm cated fr cse-How 2 z ‘ond induced efecie u Inboard vertox=—y ge | z = tao | Tip ven Zl Bs | TEL RFE toes sorte wl body vor 5 4 = ¢— 1620 Bagl ofotoc, deg ° i 20 (0) Lateral postion of vores fag falak de {© Verto poston of vortex pee Fiovmn 10.—Comparson betwedn theory aod experiment for Iteral 9 Yatra porn of verte. ee eee Fravrp 0.—Comparison between theory and experiment for lateral atio 4 trlangular wing and body combination; Ma=2.0, r/s=0.20, ‘and vertical positions of wing vortex 1.8 ey behind wing of aspect ratio 2 triangular wing and body combination; Mu=20, r/e=0.33. lation is at the wing-body juncture. Sinco the lift of the bound vortex is p,V.P'm per unit span, the value of Tq can bo estimated from tho following series of equations: Lom Lom t Laon Va aa es © To satisfy the boundary condition that the body is circular (86) tw gw tw ‘The first form of equation (36) is used for determining Dy, Since Cony p= creme rem bv) ( Ci) 7) it follows that [Karemetkres? eG CadnSe 9) ‘The problem of determining tho Interal positions of the wing vortices must be solved before the foregoing equation can be used to evaluate Tx. ‘The assumption is made that the vortices of the wing in combination are discharged at the center of vorticity of the panels of the wing alone as determined by lifting-line theory or linear theory. ‘Thi assumption is necessary because the circulation distribution is not generally known for the wing-body combination. ‘The validity of this assumption can be examined for slender swing-body combinations for which the span loading is known and from which the lateral position of the vortex ean be determined. In fact, the lateral vortex position on the basis of slender-body theory is HO (30) ‘Phis equation gives the Interal position of the vortex as a fraction of the semispan of the exposed wing panel and as a function of the redjus-semispan ratio. ‘The maximum devietion between the values given by this equation and the rwing-alone valuo of 0.786 (or m/4) is about 3 percent. ‘This result is independent of the plan form of the wing or body in front of the maximum span position since in slender-body theory the potential and, heneo, the circulation depends only on the erossflow plane under consideration. ‘For nonslender wing-body combinations the lateral posi- tion can easily be determined if the lift coefficient andLIFT AND CENTER OF PRESSURE OF WING-BODY-TAIL, COMBINATIONS the loading at the root chord are known for the wing alone. ‘The necessary equation is ae (40) In this equation (ce) is the product of the section lift coefficient at the midsection of the wing and the chord at that position. Inherent in the equation is the assumption that the maximum circulation occurs at the midsection of the wing. A series of charts hus been prepared for wings of unswept leading edges, midchord lines, and trailing edges to give the vortex location as a fraction of the wing-lone semispan and as a function of the effective aspect ratio with taper ratio as parameter. Chart 5, for subsonic speeds, is based on re- sults of DeYoung and Harper, reference 23. It is noteworthy that for low aspect ratios the lateral positions of the vortices all tend toward the slender-body value of x/4. No systematic sot of lift charts similar to those of DeYoung snd Harper is available for supersonic speeds. However, where linear- theory results are available, they were used to obtain the curves shown solid in chart 6. ‘The solid curves have been continued as dashed curves to the slonder-body value of 2/4 at zero aspect ratio for the cases in which it was felt that the extrapolation could be made safely. For the X=0 caso with no leading-edge swoop, there is a possibility that the cir- culation distribution does not have its maximum at the center line of the wing as assumed in equation (40). ‘The linear- theory solution for the load distribution for the reversed tri- angular wing is unknown for By <4. While the foregoing charts give the vortex Iateral position at the wing, the lateral position at the tal, f, is required for calculating wing-tail interference. ‘The simplest assumptions would be to set fr equal to for or fu, the asymptotic vortex Interal position, as determined from reference 11. ‘To deter- mine which of these epproximations is more accurate, both iy and fa are compared with the experimental lateral and vertical positions of the wing-tip vortex in figures 8 (a), 9 (a), and 10 (a). On the basis of this comparison and be- cause of the occurrence of the additional vortices, neither fy nor fa is superior for predicting the vortex spacing at the tail. ‘Until more data are available on vortex positions to justify a ‘moro elaborate estimate, the value of fy from charts § and 6 cor reference 24 ean be used for fp.” ‘The vertical position of the vortex at the tail can bo esti- mated by tho step-by-step caloulative procedure described in reference 25, but the process is generally toolengthy. ‘Two alternate methods aro considered. In the first, the vortex is assumed to stream backward in the free-stream direction from the wing trailing edge, ‘Tho second method, suggested by Lagerstrom and Graham, reference 11, is to ignore the effects of the image vortices, which are nearly equal and oppo- site, but to consider erossflow and the mutual effects of the external vortices. A comparison between the two positions predicted by these methods and the positions measured by Spahr and Dickey are shown in figures 8 (b), 9 (b), and 10 (b). Because of the occurrence of more than one wing vortex per panel and of body vortices, neither theoretical method ap- ears superior. ‘Therefore, it seems best to use the simpler 579 of the two methods which assumes that the vortices stream back from the trailing edge in the free-stream direction. This, assumption leads to the following equation for vortex vertical location: he=—(G—m)w sin by+[lrtEr—l—(G)w]sina (41) ‘The height is measured above the body axis and normal to it, at the center of pressure of the tail panels. Lift due to wing vortices.—For estimating the loads on the tail section, strip theory is generally applicable but the method of Alden and Schindel, reference 12, ean be applied when the necessary theoretical span loadings aro known. Tn specifying the tail load, use is made of a tail interference factor Lewin), “TB Va Gre) where (Zz) is the lift of the tail alone at anglo of attack @ The interference factor represents a nondimensional quantity useful for computing tail loads. ‘The factor i depends on the parameters Mr, (/é)r, (GB8)n (ff8)x, and (Ws)n. For a fixed body-tail configuration, the factor depends only on the vortex positions in the erossflow plane of the tail. Whether the factor 4 is ealeulated by strip theory or by the Alden-Schindel technique, several simplifying assump- tions are required regarding the wing-tail interference. ‘The first assumption is one already used in determining Kay, for Jarge aspect ratios at supersonic speeds—that the nonplanar tail section can be reduced to an equivalent planar model similar to that shown in figure 4. ‘The body is assumed to bbe flat and to act at zero angle of attack, while the tail angle of attack ar varies spanwise. ‘The second assumption is thot the lift on th tail section due to wing-tail interference is all developed by tho tail panels, even though part of it is ‘transferred to the body. In the application of strip theory to determine this lift, Lagerstrom and Van Dyke in reference 19 have shown that an exact valuo (within the realm of linear theory) is obtained for the over-all lift of the planar model if the leading edge is supersonic and the trailing edge is straight, as for a triangular wing of effective aspect ratio greater than 4. It is to be noted that the second assumption circumvents the question of whether an afterbody occurs behind the tail. Generally, the lift acting on the body is only @ small fraction of that acting on the tail section due to wing-tail interference, so that no precise consideration of the tail afterbody is usually required. Strip theory has been used to calculate a series of design charts for the estimation of 7. ‘The details of the calculations are given in Appendix B, and tho results are presented in chart 7. ‘These charts show contours of constant values of i in the crossflow plane of tho tail with the parametere Ap and (r/s)r varying from chart to chart. It is to be noted that strip theory is independent of the chord-span ratio (6s). Tn fact, strip theory represents the limiting caso of linear theory as (¢/B8)_0. ‘The charts give an immediate idea of the regions wherein wing-tail interference is most important. For triangular tails Qy—0) it is to be noted that the interference is a finite maximum when the vortex (42)580 is in the plane of the tail and slightly inboard of the tip. For all other taper ratios, however, an infinite maximum effect ocours when the vortex is at the tail tip. Strip theory is, thus, not accurate for positions of the vortex near the tail tip, except in the case of triangular wings with super- sonic lending edges, in which case it is accurate to the order of linear theory. ‘An alternate method for the determination of i is the method of Alden and Schindel, which serves as a basis for assessing the accuracy of strip theory. ‘The essential result of the method is that the lift of a lifting surface with super- sonic edges in a nonuniform flow field that varies spanwise can be evaluated to the accuracy of linear theory by the equation B= fi wanFandy 9) where wo(y) is the vertical velocity at the spanwise position y and Fy) is proportional to the span loading of tho tail at uniform angle of attack in reversed flow. Heaalet and Spreiter in reference 22 have extended the range of equation (43) to include surfaces with subsonie edges. For triangular tails with supersonic leading edges, the reversed tail is uniformly loaded so that F() is proportional to the local chord. ‘Thus, strip theory and the Alden-Schindel method give identical results for this case. Generally speaking, the Alden-Schindel technique is not suited for an analytical determination of i because, in some cases, the necessary function F(y) is not known or leads to complicated inte- grations. The Alden-Schindel method leads to results in closed form for rectangular tail and body combinations, and the calculation has been carried out in Appendix C. ‘The values of # for the vortex in the plane of a rectangular tail and for a radius-semispan ratio of 0.2 are given in Sgure 11 for four values of (o/88)z. For a value of (¢/68)r—=0 the Alden-Schindel technique and strip theory are identical ‘Pius, a comparison of the curves for other values of (¢/88)e with those for zero gives an indication of the error due to the use of strip theory for large chord-span ratios. ‘The first result is that the infinity at (jjs)=1 (for values of vs 1 -ao) \ ep ff] i ‘é gro t—+ “strip ter al tip ie wen | ae re pa ge Osu 2 en Vertex toler position, (8) Figure 11.—EWfect of chordpan ratio on lift of restangular tail due to wing vortex aa determined by Alden-Schindel technique for vortex in plane of tal; (/8)2==0.2. ference fector,/ oz REPORT 1307—-NATIONAL ADVISORY COMROTTSE YOR AERONAUTICS (¢/f8)x not equal to zoro) has been eliminated by using the Aldon-Schindel technique. ‘For vortex positions outbourd of the tail tip, the effect of (¢/8s)y is very small. However, for vortex positions inboard. of the tip, a larger offect of (c/8s)y is indicated. To obtain an idea of where the dis- crepancy due to the use of strip theory is large and whore small, @ figure has been prepared showing tho ratio of Gurion) fing 88 © measure of the error incurred in using strip theory for (¢/8s)r=0.5. This ratio is shown as a func- tion of vortex position in figuro 12. For positions of the 1 JSF I \ L- eA UT af] Oe 1012148 Verlx lotro postion, Frounn 12—Error duo to use of strip theory in estimation of loads ‘on rectangular tail section duo to wing vortices. vortex outboard of the tail tip, the error is generally very small excopt in the immediate vicinity of the tip. For posi- tions of tho wing vortex inboard of the tail tip, o maximum error of about 35 percent ean be incurred by the uso of strip theory. ‘This error decreases with increasing vortex distance from the tail. ‘The reason that larger errors are incurred for positions of the vortex inboard of the tail tip is that here tho net offect of the vortex is the small difference of large positive and negative lifts, while for outboard posi- tions the vortex induces negative lift across the entire tail. It is believed that the use of strip theory is more accurato for tapered wings than for rectangular wings sinco itis knowa to be exact for triangular wings with supersonic edges. Despite the fact that strip theory does not possess the ac- ‘curacy of Iinear theory for purposes of estimating tail londs, it has soveral decisive advantages over the linear theory (exemplified at supersonic speeds by the Alden-Schindel method). First, the necessary theoretical information is not available for using linear theory in somo cases at super- sonie speeds. Second, separate determinations would bo required. for different (c/P4)y values and for subsonic and supersonic speeds, malking tho construction of design charts extremely difficult, For these reasons and because of its great simplicity, strip theory is used in this roport for com- puting the tail interference factors except for rectangular tails at supersonic speeds. ‘Phe contribution of wing-tail interference to the lift coefficient is now derived. ‘The contribution is by definition Lan 1.87 CozentLIF? AND CENTER OF PRESSURE OF WING-BODY-TAIL COMBINATIONS with the aid of equations (88) and (42) there is obtained (Ca) Goa)» Creme ewembel i(6e re) Coons 2rAsFa—Pm) ‘Tho values of Kye and ky are obtained from chart 1, the value of i from chart.7, and the valuo of fw from chart 5 or 6. For rectangular tails at supersonic speeds the value of 7 calculated by use of the Alden-Schindel technique is recommended. (45) Pr ON WING APTERDODY DUE TO WING VORTICES In the previous work it was assumed that no change in {lateral vortex spacing occurred between the wing and tail because, for the purposes of this report, the extra work to compute the change is usually not warranted. However, if for some reason a step-by-step calculation of the vortex path is made, the lift on the wing afterbody can be estimated. ‘Tho model shown in figure 6 is used in tho estimation. ‘The lift represented by a horseshoe vortex is paVaI'n per unit span. The lift represented by the vortex system at the wing trailing edge is thus 2eVIa(fr—dw) and at the teil location is 2peVaTa(fr—ds). The net lift retained on the body between the wing and the tail is thus Lan=—2paVelalSr—on)—Ur-on) (48) With the aid of the relationships a 8) equation (46) becomes in lift coefficient form Coy — BF [See 7 49) Lagerstrom and Graham (ref, 11) have derived this same result using a different method. Generally, the change in f between wing and tail is not known unless the step-by-step solution mentioned in reference 25 is performed. In this case both the total lift and distribution of lift on tho body due to the trailing vortices is known. Howover, if only an upper bound on the valuo of Cog, i8 desired, then the value of f= can be used for fr in equation (49). SUMMARY OP LIFT COMPONENTS OF WING-ODY-TAM, COMBINATIONS ‘The soven components of the lift acting on a wing-body- tail combination are outlined as follows: 1, Lift on body nose, (w= Kv (ray (60) 2, Lift on wing in presence of body, (Awan =UEiranettmente (Coa) (1) 3. Lift on body due to wing, (Coaon=(Kxane-+ beanie) (Ore) (2) 581 4. Lift on tail in presence of body (neglecting wing vortices), (Cray Wremet rental (C.) (SE) (68) 5. Lift on body duo to tail (neglecting wing vortices), (Bom =BKaentkena] (Coo)e (SE cy 6. Lift on tail seotion due to wing vortices, 2nAg(Se—rw) (Qrn= 7, Lift on wing afterbody duo to wing vortices, Cadne [Sem] An example of the uso of these equations is presented in a subsequent numerical computation for a specific wing-body- tail combination. Chart 8, which summarizes the lift-curve slopes of wings at supersonic speeds as determined from Tinear theory, is included for use with these formulas. LONGITUDINAL CBNTER-OF-PRESSURE THEORY In the section on lft theory the differences between subsonic and supersonic speeds were given only passing attention sineo the lift theory as developed applies in the same form to both speed ranges. Tho primary affect of Mach number was manifest through the quantities (C;,)y and (Cr,)q, ‘Howover, in the center-of-pressure theory the Mach number has a direot effect on the centers of pressure of several of the lift components, and o definite distinction must be made otween the subsonic and supersonic eases for these components, Several conventions are adopted with regard to conter-of- pressure position in this report. All positions for the com- plete configuration are ultimately given in fractions of the body length behind tho most forward point of the body. In the design charts, the centers of pressure of Laon, Livny Lem, and Loe, aro given in fractions of the root chord behind the juncture of the leading edge with the body. All Iength symbols having bars over them represent center-of- pressure lengths. ‘CENTER OF PRESSURE OP HODY NOSE ‘For most purposes the center of pressure of the body nose can be estimated with sufficient accuracy by slender-body theory. The result is obtained that Va Gna) wherein Vz and Jy pre tho volume and length of that portion of the body nose forward of the shoulder. For bodies with noses of small fineness ratio or even for bodies with slender noses at high Mach numbers, some lift is carried over onto tho body behind the nose, tending to make Ty greater than tho value given by equation (67). If the lift on the nose is wn582, substantial fraction of the total lift, the effect. can be sig- nificant, In sucls cases linear theory is better then slender- body theory, although experimental values of Zy are always preferred. "In this report, the theoretical values used will be those of slender-body theory. ‘Tho centers of pressure of ogival noses as determined from slender-body theory are presented in chart 9. Angle of attack —Tho center of pressure of a triangular swing in the presence of an infinite eylindrical body as given by slender-body theory (ref. 13), in percent of the exposed wing root chord measured from the leading edgo of the wing- body juncture, is Ona gs o(ets)m G35 [ (Gr) F]48 Ce OSE) (68) An alternate method for evaluating center-of-pressure location of a triangular wing-body combination is to suppose that the exposed wings are operating in the upwash field of the body alone and then to caloulate the resultant center-of- Pressure location using strip theory. ‘The procedure to be followed is similar to that used in the lift-theory section. ‘The upflow angle duo to the body varies spanwise on the horizontal plane of symmetry as emen(145) where y is the lateral distance from the body axis. The wing is thus effectively twisted by the body-alone flow. If now the upwash angle given by equation (59) is taken into account by using strip theory, an approximate valuo of lift is given as 9) 4 Trine=F (60) of sey ‘The moment about the leading edge of the root chord is Mrunem$tuf oeiy oo It is assumed thet the center of pressure of the strip is at the midchord. Dividing moment by lift then gives for the center-of-pressure location for the wing of a triangular wing-body combination @ene 1 REPORT 1307—NATIONAL ADVISORY COMMITTEE FOR ABRONAUTICS ‘The results of equations (58) and (62) aro presented in figure 13 es e function of r/e. In addition, the value of center of pressure of the wing alone as determined by linear theory is indicated. Tt is significant that all three methods givo essentially tho samo result for the center-of-pressuro location of the wing in presenco of the body. It may bo concluded that G/o,)y for wing alone (defined as exposed swing panels joined togothor), although independent of r/e, gives a sulliciently accurate representation of Ged yin for triangular wings in presence of the body. —— sienser-body theory = Sitip theory — wing atone cet pe Eban ° 7 i Body-rasu, wing senispon roto, ‘Frovge 18.—Comparizon of theoretieal values of (ey) man for trl- ‘angular wing with no tralling-edgs sweep. If slender-body theory is applied to rectangular wings in combination, the erroneous result is obtained that all lift, and therefore the center of pressure, is at tho wing leading edge. While this result is valid for vanishing aspect ratio, it is obviously not valid in general. On the other hand, by strip theory, the center of pressure is given at the midchord and is independent of the aspect ratio. This valuo is exact only in the caso of vanishing chord and is approximately true for moderato to high aspect ratios. ‘The center-of- pressure location of wing alone as predicted by linear theory exhibits shift toward the leading edge from the midchord position with decreasing aspect ratio. Oy, Equation (63) is valid for @A>1. For BAC, negative lifting pressures due to tip effects develop on’ rearward ‘areas of tho wing, moving the center of pressure nearer the wing loading edge. ‘Thus, the wing-alone conter-of-pressure location as predicted by linear theory approaches tho value givon by strip theory for wings (in presence of body) of high aspect ratios and shows a location more in accordance with slonder-body-theory results at low aspect ratios, It is therefore concluded that for rectangular wings the center of 384-2 (63) ‘op4—3LIFT AND CENTER OF PRESSURE OP WING-BODY-TAIL COMBINATIONS pressure of the wing alone for all aspect ratios is more representative of the center of pressure of the lift on the ‘wing in presence of body than the result given by either slender-body theory or strip theory. For trapezoidal wings of no trailing-edge sweep, slender- body theory gives all the lift, and hence center of pressure, on the portions of the wing forward of the leading edgo of the tip chord, In general, however, lift is known to exist over the entire wing and the slender-body result for center- of-pressure location is too far forward at high aspect ratios. Strip theory, on the other hand, principally by not account- ing for tip effects, generally gives a center-of-pressure loca- tion too far aft of the wing leading edgo particularly at low aspect ratios. For large aspect ratios wing-alone theory is in good accord. with strip theory, and at low aspect ratios, with slender-body theory. Since strip theory is reliable only at high aspect ratios, it can be concluded that wing- alone theory is best for the entire aspect-ratio range. On the basis of the foregoing comparison of wing-alone theory with slender-body theory and strip theory for tri- angular, rectangular, and trapezoidal wings in combination with a body, it is concluded that of these three theories wing-alone theory is the best for representing the center of pressure of the exposed wing panels throughout the aspect- ratio range. Some simple charts to assist in estimating these center-of-pressure positions are now presented. For supersonic speeds, charts 10 (a), 10 (b), and 10 (c) give the variation of Gey with BA for wings of no leading-edge sweep, no midchord sweep, and no truiling-edgo sweep, respectively, for taper ratios of X=0, %, and 1. ‘The curves giving G/e)y are extrapolated to the limiting values given by slender-body theory at 6A=0, for which case slender body theory is valid. ‘The value of Gc.)y for any given wing of this family can be found by suitable interpolation. For subsonic speeds the charts of DeYoung and Harper, reference 23, can be used for estimating Eley for a wide range of aspect ratios, taper ratios, and sweep angles. ‘The results are presented in chart 11.’ Again the results have been extrapolated from values of BA=2 to the slender-body values at BA=0. Crossplotting aided in the extrapolation. ‘Tho distance from the most forward point of the body to the wing conter of pressure is Trane=ln+ (6)w@er)weme (64) 2mp-+5 Adje, __ (Bdfe,)") 1 may, B(mp-+1¥" Sonp+iy am | 14 Sme-+1) p+ edmp +1) cos? Bae on (+9i)- Lae] 583 Wing-incidonce angle—No goneral method for estimating Gedy eSists, but specialized results are available for rectangular wing and body combinations for which BA>2 or for slender triangular wing and body combinations. For the rectangular wing and body combinations, values of Ger) based on linear theory obtained from reference 3 aro presented in chart 12, The values of Glens a0 lower than the wing-alone (r/e=0) values by a few percent of the root chord. ‘The results for slender trianguler-wing and body combinations as determined from slender-body theory. in Appendix A are shown in chart 13. The deviation of Gloria, from the wing-alone value of % is only @ fractiona] percent of the root chord. For the combination to which they apply, the results of charts 12 and 13 are to be used. For other combinations, (2/c,)y- provides a good approxima tion to Gedy, Until more accurate values are available, ‘Tho center-of-pressure position with reference to the body Tor (65) lint Cw Elen (CENTER OF PRESSURE ON BODY DUE TO WING ‘Tho conter of pressure acting on the body duo to the wing is determined by different methods, depending on whether subsonic or supersonic flow is considered. ‘The assumption is made that the center of pressure of the lift transferred from. the wing to the body is not sensitive to whether the lift is developed by angle of attack or by wing deflection. ‘Then there is no appreciable difference between Gc.) gqrye and Ge) yy» And these two cases are not treated separately. Supersonic flow.—For the supersonic case the planar modal of figure 4 is used. This is the samo model that was used for the determination of tho lift on the body in the presence of the wing. The moment of the lift (eq. (23)) carried onto the body by a wing with a supersonic leading edge is £ Bt bmn 4qcy6m aes Maen ele flan fg coor rapt (60) in terms of the coordinate system of figure 4(b). ‘This result, doubled to account for the lift of two half-wings, gives ile (Bdfe,)* _ me [Hata imi mp Vn (67)